EP0649696B1

EP0649696B1 - Method for classifying discharge machining wave patterns, and method for preventing arcs based on the classification of the discharge machining wave in discharge machining apparatus

Info

Publication number: EP0649696B1
Application number: EP93402601A
Authority: EP
Inventors: Seo Seok-Yong; Kim Byung-Guk; O. Seung-Yeop; Kim Doo-Won; Jeong Ju-Yong
Original assignee: Korea Atomic Energy Research Institute KAERI
Current assignee: Korea Atomic Energy Research Institute KAERI
Priority date: 1993-10-22
Filing date: 1993-10-22
Publication date: 1997-10-08
Anticipated expiration: 2013-10-22
Also published as: EP0649696A1; US5453593A

Description

Field of the invention

The present invention relates to a method for classifying electro-discharge machining wave patterns by utilizing a delay circuit, and also relates to a method for preventing arcs based on the classification of the electro-discharge machining wave patterns in an electro-discharge machining apparatus. That is, the patterns of the voltage or current waves which are generated during an electro-discharge machining (EDM) are classified by utilizing a delay circuit. According to the classifying method, the generation of harmful or non-harmful waves is monitored.

Background of the invention

Conventionally, there have been disclosed a number of the wave pattern classifying methods for the purpose of monitoring the generations of harmful or non-harmful waves.
As the principal approaches for monitoring the state of the electro-discharge machining, there are methods such as: analyzing the radio frequency generated during the discharge sparking; measuring and analyzing the electric resistance between the electrode and the object to be machine; and analyzing the discharge wave patterns. Among them, the most preferred was the method of analyzing the wave patterns.
The studies on the analysis of the electro-discharge machining wave patterns which have been carried out so far can be summarized as follows.
That is, the following persons studied on the electro-discharge machining wave patterns: R. Snoeys (CIRP vol 24, 1974), S Bhattacharyya (ASME J. of Eng. Ind., 1980), M. Otto (ISEM7, 1983), A. Endel (ISEM7, 1983), S. Pandit (ASME J. of Eng. Ind., 1984), D. Dauw (CIRP vol 35, 1986), S. Pandit, ASME J. of Eng. Ind., 1987), and C. Cogun (ASME PED vol 34, 1988).
The contents of the conventional studies are different from the first object of the present invention in one point. That is, there is a difference in the sampling method. In the conventional studies, there is never found the sampling method of the present invention. In order to clarify this point, the known facts about the discharge wave patterns which occur during an electro-discharge machining operation will be briefly described.
Generally, during a transistorized pulsed electro-discharge machining (to be called "electro-discharge machining" below), 5 types of voltage or current wave patterns ( $voltage = current x resistance$
) are generated in a random manner. That is, waves of the normal or effective discharge, arc, short, open and semi-normal discharge (the last one being of the present invention) are generated (refer to Figure 1). Of them, only the normal discharge and the semi-normal discharge waves are effective ones. The open wave is not desirable because it lowers the machining efficiency, while the short and arc can give damages to both the machining apparatus and to the object to be machined, if they continue for a long time. Accordingly, if the 5 wave patterns can be classified at a high speed, it can contribute to monitoring the electro-discharge machining.
The conventional wave pattern classifying method which is most similar to the first object of the present invention will be described below. As shown in Figure 2-2, the wave patterns are classified by comparing high and low reference levels with the levels of waves which are produced during the electro-discharge machining.
However, there is a difficulty in classifying the 4 types of the waves, if only the comparison of the voltage levels is used. Therefore, another reference level or another sampling method has to be introduced. In this case, the sampling is carried out by distinguishing the discharge into the initial period and last period.
Referring to Figure 2-2, the portions indicated by (1) are the initial period of the discharge, while the portions indicated by (2) are the last period. That is, a sampling is carried out within a certain time ta after the starting of the discharge, and another sampling is carried out after the first sampling. As a result, 4 types of waves can be classified with the condition of the logic diagram of Figure 2-2. Figure 2-1 illustrates the same details in a schematic view. In this case, the crucial disadvantage is that the semi-normal wave cannot be detected. If the time ta is taken short, the distinction between the normal or effective discharge and the arc is clear, but the classification for the semi-normal and open waves becomes impossible.
It should be understood that the semi-normal waves greatly contribute to the machining. Meanwhile, if the time ta is taken long, the classification for the normal or effective discharge waves and the arc waves becomes impossible.
Conventionally, a number of methods have been disclosed for the purpose of monitoring the generation of harmful arcs. The discrimination as to whether there are arcs are made based on the increase of the average current or the drop of the voltage across the discharge electrode, thereby preparing against the generation of arcs. For example, measures such as off-time and the like are taken, with the result that effective results are gained.
However, such a conventional method is different from a second object of the present invention in two points. That is, a discrimination is carried out for every discharge pulse, and then, a proper measure is taken before the completion of the pulse. In other words, the measurement, classification and adjustment are almost simultaneously carried out for every pulse.
GB-A-2 115 585 discloses an electric discharge machining apparatus, which automatically optimizes the machining conditions, in order to obtain the maximum effective discharge rate. This documents suggests distinguishing digitally abnormal discharges from effective pulses. For this purpose, it is suggested to use timing pulses occurring at the beginning of each pulse, at the end of each pulse, and immediately after the end of each pulse. The proposed apparatus uses not only simple logic components, but also optimization controllers and controllers.

Summary of the invention.

The invention relates to a method for classifying electro-discharge machining (EDM) wave patterns, as defined in claim 1. It also relates to a method for preventing arcs or shorts by classifying electro-discharge machining (EDM) wave patterns, as defined in claim 6. Advantageous embodiments of the invention are disclosed in the dependent claims.
A first object of the present invention is to provide a method in which two short samplings are carried out within a short time period ta for every discharge wave, in such a manner that one sampling is carried out at the initial stage of the wave, and another sampling is carried out at the last stage of the wave.
A second object of the present invention is to provide a method for classifying the patterns of the voltage and current waves which are generated during an electro-discharge machining operation, and to provide a method for preventing arcs based on the results of the classification.
Now the first object of the present invention will be described in more detail.
In achieving the first object of the present invention, the current wave patterns are classified unlike the conventional methods.
First the generation of the discharge wave will be described.
The electro-discharge machining is carried out in such a manner that a dielectric (made of hydrocarbons such as light oil) is inserted between an electrode and an object to be machined to form a gap between them, and dc pulse voltages are supplied to the gap, so that sparking should occur under the proper condition. Owing to such an energy, a part of the object to be machined is evaporated assisted by the great vaporizing pressure of the dielectric medium. As a result, a crater is formed on the object to be machined, and the average accumulation of the craters results in the machining of the object. Under this condition, the patterns of the voltage and current waves can be sampled. However, depending on the conditions, there is no guarantee that sparking will necessarily occur every time. For example, if the inter-electrode gap is too large, open waves will be generated, and an actual machining will not occur. On the other hand, if the inter-electrode gap is too small, short-circuit waves will be generated, and also machining will not be done. Further, a portion may have too small resistance compared with the rest of the portions due to the turbidness of the dielectric medium or other reasons. In this case, the discharge may be focused on the spot where the resistance is smaller, and this is called arc. Generally, the electro-discharge machining apparatus should be designed such that the apparatus should generate the normal or effective discharge or the semi-normal discharge. However, practically, several or several score percents of ineffective or harmful waves are generated.
The primary purpose of the classification is to monitor to decide good or bad. Further, the obtained information can be used for self-controls.
Regarding the classification by potential, a variable voltage which is formed by adjusting the received discharge waves is compared with a reference voltage which is set in advance by a circuit.
Of the classification methods, the classification by a logic circuit will now be described in detail. The classification is made by combining the methods by potential and by the logic circuit.
The determination of the timing for the sampling is the most crucial point of the first object of the present invention. Regardless of the types, the discharge waves are repeating waves in which a cycle consists of an on-time and an off-time. In the first object of the present invention, the variables are sampled only at the initial stage of the on-time and immediately before the off-time. During the rest of the time, all the variations are disregarded. Particularly, it is the most important feature of the first object of the present invention to make a sampling immediately before the termination. Particularly, it proves the superiority of the first object of the present invention that the above function can be carried out without receiving impediments even under the random variation of the on-time and off-time.

Brief description of the drawings

The above object and other advantages of the present invention will become more apparent by describing in detail the preferred embodiment of the present invention with reference to the attached drawings in which:

Figure 1 is a schematic view of the discharge waves for illustrating the method of classifying the electro-discharge machining wave patterns utilizing a delay circuit;
Figure 2-1 is a logic chart for illustrating the principle of the known discharge wave classification method for showing the classification method based on a delay circuit;
Figure 2-2 is a schematic view of the wave patterns of Figure 2-1;
Figure 3 is a logic diagram illustrating the classification of the electro-discharge machining wave patterns based on a delay circuit;
Figure 4 illustrates the classification of the electro-discharge wave patterns utilizing a delay circuit, in which, even if the on-off magnitudes of the reference clock for deciding the length of the pulse of the discharge wave is varied in a random manner, the short monitoring signals which are formed by an AND logic combination of the signals delayed for a certain period of time are generated at the same position relative to the reference pulses;
Figure 5-1 illustrates an example of the classification of electro-discharge machining wave patterns using a delay circuit, in which the signals for arcs and the arc waves detected by the method of the present invention are shown;
Figure 5-2 illustrates the detected open waves similarly to Figure 5-1;
Figure 5-3 illustrates detected short circuit wave similarly to Figure 5-1;
Figure 6 illustrates schematic discharge wave patterns for showing the method of preventing arcs based on the classification of the discharge wave patterns, the upper portion of the drawing showing the discharge current waves, and the lower portion showing the pulse of an electro-discharge measuring clock (EDM clock);
Figure 7 illustrates the time position (positioned always at the start of the on-time) relative to the EDM clock for the sampling pulse for showing the arc preventing method based on the classification of discharge wave patterns, the upper portion of the drawing showing the EDM clock, and the lower portion showing the monitoring pulses;
Figure 8-1 is a logic chart illustrating the principle of the arc preventing method based on the classification of the discharge wave patterns;
Figure 8-2 illustrates the same contents as that of Figure 8-1;
Figure 9-1 illustrates the result of an example for illustrating the arc preventing method based on the classification of the discharge wave patterns, the upper portion of the drawing showing the discharge current, and the lower portion showing the arc generation signals based on the monitoring pulse; and
Figure 9-2 illustrates a part of the effect for showing the arc preventing method based on the classification of the discharge wave patterns, the upper portion of the drawing showing the discharge current, and the lower portion showing the EDM clock.

Description of the preferred embodiment

In the logic chart of Figure 3, S represents discharge current variable (to be called input signal below), H a high reference potential set in advance, and L a low reference potential. As to the current wave patterns, only 5 typical ones are shown for the sake of the convenience of description. Starting from the left side of the drawing, a normal or effective wave, an open wave, an arc wave, a short wave and a semi-normal wave are shown. It should be noted that the shapes of the waves are opposite to those of the voltage waves.
It also should be noted that the normal or effective wave and the semi-normal wave are similar to each other. The waves which are capable of machining are the normal or effective wave, the semi-normal wave and the non-definite curve (oscillation) of a part of the arc wave. Therefore it should be understood that it is incorrect and erroneous to classify the semi-normal wave as an open wave as in the conventional method.
In Figure 3, 〈1〉 indicates a logic signals for the case where the high potential is higher than the input signals. 〈2〉 indicates a logic signals for the case where the input signals are higher than the low potential level. 〈3〉 indicates an AND logic signal for 〈1〉 and 〈2〉.
〈4〉 and 〈5〉 indicate inverted signals of 〈1〉 and 〈2〉 respectively. 〈9〉 indicates a signal of a main clock, and here, the magnitudes of the on and off are arbitrarily varied in accordance with the operating status of the discharge apparatus.
〈10〉 indicates an inverted signal formed by delaying the signal 〈9〉. In this case, the amount of delay is set in advance, and usually, the amount of delay is 10 micro seconds.
〈6〉 indicates an AND signal for 〈10〉 and 〈3〉. 〈7〉 indicates an AND logic signals for 〈10〉 and 〈4〉, while 〈8〉 indicates an AND logic signal for 〈10〉 and 〈5〉.
〈11〉 indicates a single shot signal generated by the rise of the signal 〈10〉, and it belongs only to the signal 〈10〉. This signal is used as a monitoring signal. 〈12〉 indicates an AND logic signal for 〈9〉 and 〈10〉, and this signal is used as a monitoring signal. 〈13〉 indicates an AND logic signal for 〈7〉 and 〈11〉. This signal is generated by the short wave, and therefore, this signal can be used for detecting short waves.
〈14〉 indicates an AND logic signal for 〈8〉 and 〈12〉. This signal is generated under the open state, and therefore, this signal can be used for detecting open waves.
〈15〉 indicates an AND logic signal for 〈6〉 and 〈11〉. This signal is generated only during an arc, and therefore, this signal can be used for detecting arc waves.
The normal or effective wave and semi-normal wave do not have to be detected.
A special feature is that two samplings are carried out during one cycle by utilizing the signals 〈11〉 and 〈12〉, and that the signal 〈12〉 is positioned immediately before the off of the signal 〈10〉 which is the discharge gate signal. Particularly, as shown in Figure 4, even if the on and off are randomly varied, the sampling can be carried out immediately before the off of the signal 〈10〉 (〈B〉 in Figure 4). This advantage is not seen in any other conventional methods.
Now an actual example for the first object of the present invention will be described.

〈Example〉

An electro-discharge machining apparatus was driven by square waves in which the duration of one cycle was 480 micro seconds, and the on-time was 85%. The current signals of this apparatus was supplied to a wave pattern analyzing instrument, in order to see if the characteristic signals corresponding to the open wave, arc wave and short wave are outputted. The detector used was an ordinary oscilloscope, and the plotter used was Type A3.
A channel 2 was mode to receive discharge current waves, while a channel 1 was made to receive in a sequential manner the characteristic signals corresponding to the above three waves. If all the signals are to be received simultaneously, a 4-channel instrument should be required.
The experiment results are as shown in Figures 5-1, 5-2 and 5-3.
In Figure 5-1, the electro-discharge machining current signals and the detected signals corresponding to the arc signal pulses appear at the same time position. From this fact, it can be known that the arc signals and the arc detection signals well correspond to each other.
Figure 5-2 illustrates a fact similar to Figure 5-1, and it can be known that the open wave detection signals well correspond to the open waves. (Refer to the claim).
Figure 5-3 illustrates a fact similar to Figure 5-1, and it can be known that the short detection signals well correspond to the short waves.
The signals 〈13〉, 〈14〉 and 〈15〉 can be used for monitoring the state of the electro-discharge machining apparatus, and provide a decisive information for the function of self-control.
Now descriptions will be made in connection with the second object of the present invention.
There are a number of methods for monitoring the state of the electro-discharge machining such as: analyzing the radio frequency generated during sparking; measuring and analyzing the electric resistance between the electrode and the object to be machined; and analyzing the discharge waves. Among these methods, the most preferred is the method of analyzing the discharge waves.
The researchers who have contributed to the analysis of the discharge waves are as follows: R. Snoeys (CIRP vol 24, 1974), S. Bhattacharyya (ASME J. of Eng. Ind., 1980), M. Otto (ISEM7, 1983), A. Endel (ISEM7, 1983), S. Pandit (ASME J. of Eng. Ind., 1984), D. Dauw (CIRP vol 35, 1986), S. pandit (ASME J. of Eng. Ind., 1987), and C. Cogun (ASME PED vol 34, 1988).
The significant difference between the second object of the present invention and the known studies is that the present invention uses the wave classification result for preventing arcs. In order to assist understanding on this matter, the known facts on the discharge waves which are generated during an electro-discharge machining will be briefly described below.
Generally, when carrying out a transistorized pulsed electro-discharge machining (to be called discharge machining below), there are generated 4 types (or 5 or 6 types in some researchers) of voltage or current waves ( $voltage = current x resistance$
). That is, these waves are normal or effective discharge (or spark), arc, short and open waves (refer to Figure 6), and, among them, the normal or effective discharge is the working component. The open waves are undesirable because they lower the machining efficiency, while the short and arc waves are also undesirable because they may give damages to both the machining apparatus and to the object to be machined, if they continue for a long time. Therefore, if it is possible to classify the above mentioned 4 types of waves, it would contribute to monitoring the discharge machining.
According to the second object of the present invention, the above waves are classified, and, if an arc wave or a short wave is generated (an arc signal is always positioned at the leading end of each cycle), the power supply is disconnected during the relevant cycle or during a small fraction of the relevant cycle.
This second object of the present invention will be described in detail below.
In connection with the second object of the present invention, the current waves are utilized, and the wave patterns are classified, while adverse effects due to the arc or short are prevented (in the case of the voltage waves, the explanation will be same).
The principal purpose of the classification of the discharge wave patterns is to monitor the discharge machining, thereby distinguishing the good or bad quality of the machining. Further, this information can be used for a self-control.
According to the second object of the present invention, by utilizing this information, the power supply is withheld upon the occurrence of an arc or short, so that the discharge machining apparatus and the object to be machined should be protected from damages.
The classification is carried out in the manner described below.
Within a predetermined sampling period, the discharge waves are received, and then, the variable which is adjusted to a proper voltage range is compared with a reference voltage which pre-set by the circuit. The reference voltage consists of a high level and a low level. The level setting range is such that the high level corresponds to 95 to 99 % of the peak value of the current wave, and the low level corresponds to 25 to 90%.
The sampling time is determined as follows. That is, regardless of the types of the waves, the cycle of the discharge wave repeats, and the cycle consists of on-time and off-time. According to the second object of the present invention, the sampling of the variable is carried out only at the start of the on-time (refer to Figure 7). Further, monitoring pulses for the sampling are generated simultaneously with the discharge on-off signals by utilizing a main clock timer of the computer, and therefore, even if the lengths of the on-off signals are varied in a random manner, the variable is generated always at the same on-time position.
Regarding the logic chart, the relevant drawing illustrates by what logic structure the arc waves are detected from the discharge waves, and also illustrates by what logic structure the discharge power source is disconnected upon the generation of an arc wave. There are two methods of blocking the discharge power. That is, one is that in which the blocking period is fixed, and the other is that in which the blocking period is variable.
In Figures 8-1 and 8-2, W indicates discharge current variable (to be called "input signal" below), H indicates a pre-set high reference level, i.e., 95 to 99 % of the peak value of the input signal, and L indicates a pre-set low reference level, i.e., 25 to 95 % of the peak value of the input signal. The current wave patterns are classified into 5 typical patterns for the sake of the convenience of the descriptions. Further, the portions which have no direct relationship to the present invention are omitted.
Referring to Figure 8-1, the curve shows from the left the normal wave, the open wave, the arc wave, the short wave and the semi-normal wave. The dotted lines show the shape of the original input signals, and the solid lines show the modified shape after the blocking of the power. That is, when an arc or short occurs, the wave pattern is modified as shown by the solid lines, resulting in that the arc and short are blocked.
It should be noted that the normal wave and the semi-normal wave are similar to each other. The waves which do actually machining are the normal wave, the semi-normal wave and a part of the arc wave.
Here, in the case of a fixed blocking, 〈2〉 in Figure 8 indicates a logic signal for the case where the input signal is higher than the low reference level. 〈14〉 indicates the signal of EDM clock (electro-discharge machining clock), and here, there is the case that the magnitudes of the on-off are arbitrarily varied in accordance with the operating state of the discharge apparatus.
〈6〉 indicates a sampling pulse which is continuously generated from the sampling circuit, and which appears always at the start of the signal 〈14〉. (Mostly Timing IC 8254 is used).
〈9〉 indicates an AND logic signal for 〈2〉 and 〈6〉, and this signal proves the generation of an arc.
〈17〉 indicates an inverted output Q or $\bar{Q}$
which is inputted into a D-flipflop which uses the signals 〈6〉, the inverted output Q or $\bar{Q}$
being formed from the signal 〈2〉.
〈15〉 indicates an AND logic signal for 〈14〉 and 〈17〉, and shows that, during the occurrence of an arc, the whole portion of the on-time becomes equivalent to the low level. That is, if the discharge power is controlled by using the signal 〈15〉, then the discharge power is automatically blocked during the occurrence of an arc.
In the case of a variable blocking, 〈2〉 indicates a logic signal for the case where the input signal is higher than the low reference level. 〈14〉 indicates an EDM clock, and here, the magnitudes of the on and off may be arbitrarily varied in accordance with the operating state of the discharge apparatus.
〈6〉 indicates a sampling pulse which is continuously generated from the sampling circuit, and this pulse 〈6〉 is positioned always at the start of the signal 〈14〉. (Mostly Timing IC 8254 is used).
〈9〉 indicates an AND logic signal for 〈2〉 and 〈6〉, and this signal proves the occurrence of an arc.
〈23〉 indicates a main clock of the electronic calculator.
〈17〉 indicates a signal which is obtained by connecting the signal 〈9〉 to the gate of a down counter (e.g., IC 8254) using the clock 〈23〉. The range of the down-count is set by a program of a computer.
〈15〉 is an AND logic signal for 〈14〉 and 〈17〉, and, in this signal, it is shown that the whole portion of the on-time is shifted to a low level during the occurrence of arc. That is, if the discharge power is controlled by using the signal 〈15〉, then the discharge power is automatically blocked during the occurrence of an arc.
Here, the characteristic feature is that the input signal is sampled at the start of the on-time of every cycle by using the signal 〈6〉, and that the signal 〈15〉 is accurately put to an off-state upon the occurrence of an arc.
Now an actual example for the second object of the present invention will be described below.

〈Example〉

A discharge machining apparatus was used, and the apparatus was driven by square waves in which one cycle was 50 micro seconds, and the on-time was 30 micro seconds. Then the discharge current signals of the discharge machining apparatus was supplied to an analyzing instrument. Thus the characteristic monitoring signals corresponding to the arc waves were obtained. Under this condition, in order to confirm as to whether the discharge current wave is in an off state, the above two signals were simultaneously recorded on an oscilloscope. Figure 9-1 shows the experiment results. As shown in Figure 9-2, in the cycle in which an arc monitoring signal clearly exists, it is apparent that a part of the discharge current is not being supplied. The reason for the non-uniformness of the blocking period is that breakdown delay occurs due to the dielectric.
To state the effect of the second object of the present invention, even under the adverse condition as in Figure 9-2 in which the arcs are continuously generated, the discharge machining could be performed in a perfect manner.
If the signal 〈15〉 is supplied to the gate of a power transistor, the intended purpose can be achieved. That is, the arc or short signals which are outputted from the discharge wave analyzer is modified by the AND-logic-combining the EDM clock to the signal 〈15〉 by the function of the above described logic circuit. Thus, when an arc or short occurs, the discharge becomes impossible in the relevant cycle.
According to the present invention as described above, a discharge machining apparatus can be made in which damages due to an arc or short do not occur. Therefore, not only the damages due to the arc are eliminated, but also, the off-time can be reduced to an extreme degree, so that the machining efficiency can be improved.

Claims

A method for classifying electro-discharge machining (EDM) wave patterns, comprising the steps of
comparing the voltage or current wave pattern (S) generated during the EDM operation with a reference level consisting of a high level and low level (L, H) to produce logic signals (1, 2);

delaying and inverting (10) a main clock (9) signal of the EDM machine to produce a discharge gate signal (10);

providing a first monitoring signal (11) at the start of an on-time of said discharge gate signal (10),

subject the main clock signal (9) to an AND logic combination with the discharge gate signal (10) to form a second monitoring signal (12);

sampling said logic signals (1, 2) with the first monitoring signal (11) and with a second monitoring signal (12), an arc or short wave pattern being classified by sampling at the first monitoring signal (11).
A method according to claim 1, wherein an arc or short wave pattern is classified when a current wave pattern (S) is higher than said low level (L) at the first monitoring signal (11) at the start of an on-time of said discharge gate signal (10).
A method according to claim 1 or 2, wherein an arc wave pattern is classified when a current wave pattern (S) is higher than said low level (L) and lower than said high level (H) at the first monitoring signal (11) at the start of an on-time of said discharge gate signal (10).
A method according to claim 1, 2 or 3, wherein an open wave pattern is classified when a current wave pattern (S) is not lower than said high level (H) at the first monitoring signal (11) at the start of an on-time of said discharge gate signal (10).
A method according to one of claims 1 to 4, wherein an short wave pattern is classified when said wave pattern (S) is not higher than said low level (L) at the second monitoring signal (12).
A method for preventing arcs or shorts by classifying electro-discharge machining (EDM) wave patterns, comprising the steps of
comparing the voltage or current waves (W) generated during a transistorized pulsed EDM operation with a reference potential (L) during a high level of a monitoring pulse (6 fig. 8-1) at the start of an on-time, said reference potential corresponding to 25 to 95 % of the peak value of said current wave, to obtain signals corresponding to the arcs or shorts (9),

using the said signals (9) to withhold the power supply to prevent the arcs or shorts.
A method according to claim 6, wherein the step of using comprises the steps of:
inverting and flip-floping (17) said signals by using said monitoring pulse as the clock

subject said inverted and flip-flopped signals (17) to an AND-logic combination with an EDM clock (14);

supply the AND-logic combined signal (15) to the gate of a discharge controlling power transistor to prevent arcs or shorts.
A method according to claim 6, wherein the step of using comprises the steps of:
trigger the gate of a down counter with said signals (9);

subject the output (17) of said down counter to an AND-logic combination with an EDM clock (14);

supply the AND-logic combined signal (15) to the gate of a discharge controlling power transistor to prevent arcs or shorts.
A method according to claim 6, 7 or 8, wherein the monitoring pulse are generated simultaneously with the discharge on-off signals of the EDM machine.